Overview of AI Part 2

CS123, Intro to AI

Table of Contents

Introduction

Things to Do This Week

• Join the Discord server so you can communicate with your team.

• Introduce yourself survey—complete it by Friday.

• We'll choose teams next Monday.

• Read Elements of AI Ch. 1 and do the exercises by Sunday

• Answer "Your turn" questions in the forum by Sunday
  This is mainly for online students or anyone who wasn't able to participate in the live classroom discussion.

• Take the quiz over the lectures by Sunday

Dangers of AI

Sure, here's an outline of the dangers of AI, including ethical, social, privacy, and philosophical issues:

1. Ethical Issues

   • Bias: AI systems can unintentionally perpetuate and amplify existing biases.

   • Data Exploitation: AI's ability to gather and analyze massive quantities of data can lead to exploitation of personal information.

   • Decision Accountability: With generative AI in particular, there is a lack of transparency in AI decision-making processes, leading to accountability issues.

   • Environmental Impact: The energy consumption of AI systems can have significant environmental implications.

2. Social Issues

   • Job Automation: AI could lead to high unemployment rates due to automation of work.

   • Socio-economic Inequality: The rise in unemployment and AI bias can create socio-economic inequality.

   • Abuse of Data and Loss of Control: There's a risk of misuse of data collected by AI systems.

   • Financial Instability: Rapid advancements in AI could lead to financial instability.

3. Privacy Issues

   • Data Persistence: Data can exist longer than the human subjects that created it.

   • Data Repurposing: Data can be used beyond their originally intended purpose.

   • Covert Metadata Collection Practices: AI systems can collect metadata without users' knowledge.

   • Limited Built-In Security Features: AI models often have limited security features, increasing the risk of data breaches.

4. Philosophical Issues of Personhood and Humanity

   • AI Consciousness and Agency: There's a debate on whether AI can exhibit personhood, including consciousness and agency.

   • Moral and Legal Status of AI: If AI systems can achieve consciousness and personhood, what will the moral and legal rights of these systems be?

   • Impact on Humanness: If AI systems appear to be human and we treat them as human, even though they are machines, will people begin to think of each other as machines? Will our concept of personhood be diminished? See "You are not a parrot" by Emily Bender.

Your turn: What are some concerns you have about AI?

Reference

• Elements of AI—The University of Helsinki and MinnaLearn

• The Turing Test—Wikipedia

• You are Not a Parrot—Emily Bender, New York magazine

• Climbing towards NLU: On Meaning, Form and Understanding in the Age of Data—Bender and Koller

  • Video and slide presentation by the authors on SlidesLive

• To Dissect an Octopus: Making Sense of the Form/Meaning Debate—Juan Michael

Intro to AI Lecture Notes by Brian Bird, written in 2024, are licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.